Joint Manifold Learning for Integrated Analysis of Single-Cell Data

To develop statistical methods to integrate multiple single-cell genomics datasets, with a particular focus on applications in early development.

This collaborative project built off their earlier publication to optimize a new computational approach to perform batch effect correction in single-cell RNA-seq data. Their approach of using Mutual Nearest Neighbors improved speed, efficiency, and accuracy relative to prior implementations. While this work remains in progress, many of these improvements were incorporated into the widely used R package “scran,” and MOFA+, a statistical framework for the comprehensive and scalable integration of single-cell multi-modal data.